i think the inpainting for general audio is ok. but for speech inpainting, there is still issue.
check this command for tts inpainting, not working properly
audioldm2 -t "A male reporter is speaking" --transcription "I have a dream" --file_path 00008-00039.wav --mode sr_ip
i think the inpainting for general audio is ok. but for speech inpainting, there is still issue.
check this command for tts inpainting, not working properly audioldm2 -t "A male reporter is speaking" --transcription "I have a dream" --file_path 00008-00039.wav --mode sr_ip